3 1. Introduction ngicering Stare Vol 2, No 4-8, pp 261-270, 1958 (© 1991 Pubs by Elsevier Scieace Lad All sigs reserved. Pit in Grek Beta PIT: $0141-0296(97)00032-1 ‘141-029658 $1900 + 000, Lessons from damage to steel buildings during the Northridge earthquake Stephen A. Mahin Deparment of Civil and Environmental Engineering, University of California, Berkeley, €4.94720, USA One of the important overall surprises of the Northridge earthquake of 17 January 1994, was the widespread and unanticipated brittle fractures in welded steel beam-column connections. The econ- omy, versatility and presupposed high plastic deformation capacity of welded steel moment-resisting frame (WSMF) buildings led to their common usage in Los Angeles and elsewhere in the US. No casualties or complete collapses occurred during the Northridge earthquake as a result of these connection failures, and WSMF buildings in areas of moderate shaking were not damaged at all. However, a wide spectrum of brittle connection damage did occur, ranging from minor cracking observable only by nondestructive testing to completely severed columns. This paper reviews the per- formance of steel buildings during the Northridge earthquake, and examines some of the results of studies undertaken as part of a project initiated by US Federal Emergency Management Agency (FEMA) to reduce the earthquake hazards posed by steel moment- resisting frame buildings. The objective of this project is to develop and verify reliable and cost-effective methods for the inspection, evaluation, repair, and rehabilitation of existing steel frame build- ings and the construction of new ones. The project is conducted by the SAC Joint Venture, consisting of the Structural Engineers Association of California, the Applied Technology Council and the California Universities for Research in Earthquake Engineering. © 1997 Published by Elsevier Science Ltd. more floors failed, or significant permanent lateral displace- ments occurred. In a couple of cases, the damage was so Every earthquake provides new lessons for the earthquake ‘engineering profession. The widespread and unanticipated damage to welded steel moment resisting frame systems ‘was one of the major overall lessons of the Northridge earthquake. While no life loss was associated with this damage, the brittle nature of the fractures detected in ‘numerous welded steel bear to column connections, essen: tially invalidated historic design approaches and code pro: ‘The most commonly observed damage occurred in or near the welded joint of a girder bottom flange to the sup- porting column flange; complete brittle fractures occurred jn many cases, The damage was so severe in some build- ings that all of the moment resisting connections at one or severe that buildings were demolished or evacuated follow ing the earthquake. The incidence of damage to steel moment resisting frames was sufficiently high in regions of strong motion to cause the wide-spread concern of struc- tural engineers and building officals, as well as owners, lenders and insurers. No comprehensive tabulation is yet available for all of the steel buildings damaged in the Northridge earthquake. ‘Thus far, more than 150 damaged buildings have been identified, including hospitals and other health care facili- ties, government, civic and private offices, cultural and edu: cational facilities, residential structures, and commercial ‘and industrial buildings. The damage occurred in new as well as old buildings, in tall as well as in short structures.262 Damage to steel buildings during the Northridge earthquake: S. A. Mahin While inadequate workmanship was believed to play a role in the damage observed in some structures, most damaged buildings are believed 10 be constructed consistent. with ‘moder codes and standards of practice. The effect of these ‘observations has been a loss of confidence in the procedures used in the past to design and construct welded connections in steel moment frames, and a concern that structures incorporating these connections may not be sufficiently safe, A panicularly disconcerting aspect of this damage is that it often occurred without accompanying distress to architec- tural finishes and cladding. As a result, reconnaissance reports immediately following the Northridge earthquake often cited the apparent excellent behavior of steel frame buildings. However, severe damage found in buildings under construction ‘at the time of the earthquake, and detailed investigations of WSMF buildings which suffered increasing amounts of damage during aftershocks, quickly identified the true performance. However, the absence of ‘outward signs of damage has made it dificult to determine ‘which buildings should be inspected to see if they have sustained structural damage during the Northridge earth- quake, ‘Current professional judgment is that the historic prac- tices used for the design and construction of WSMF con- nections do not provide adequate reliability and safety, and should not continue to be used in the construction of new buildings intended to resist earthquake ground shaking through inelastic behavior. As a consequence, pre-qualified connection details and design methods contained in the major US building codes have been rescinded, and emerg- ency code provisions stipulate that new designs must be substantiated by testing or test-backed calculations. None- theless, several fundamental questions must be answered in order to develop effective and economical design pro- cedures and construction standards, and to restore public and professional confidence in this form of construction. These questions include: # What happened to WSMF buildings during the Northridge earthquake? What caused the observed damages? * How to identify WSMF buildings that may have sus- tained damage? ‘How safe are damaged WSMF buildings and do they need to be repaired? * How can damaged buildings be reliably repaired and/or upgraded? + How to design and construct new buildings so they will not sustain similar damage in future earthquakes? * Can the vulnerability of existing WSMF buildings 1 future earthquakes be reliably determined and mitigated through effective rchabiltation procedures? # What are the economic, social and politcal impacts of any required changes in design and construction prac tices? Answering these questions involves the consideration of ‘many complex technical, professional and economic issues including metallurgy, welding, fracture mechanics, connec tion behavior, system performance, and practices related to design, fabrication, erection and inspection. Unfortunately, ‘our current knowledge of many of these issues relating to seismic performance of WSMF structures is inadequate. To resolve this situation, the US Federal Emergency Manage- ‘ment Agency has initiated a 5 year program to develop and verify reliable and costeffective methods for the inspec- tion, evaluation, repair, rehabilitation and construction of steel frame structures, This paper reviews the scope of the efforts being under- taken under FEMA sponsorship, focusing primarily on issues related to evaluation and repair of existing WSMF structures, and highlights some of the results and rec- ‘ommendations developed to date 2. Program to reduce earthquake hazards in steel moment frame structures ‘A. coordinated, problem-focused program of research, investigation and professional development has. begun under FEMA-sponsorship in order to develop and validate reliable and cost-effective seismic-resistant design and rehabilitation procedures for WSMP structures. This pro- gram is being managed and administered by the SAC Joint Venture. This joint venture consists of three not-for profit professional and educational organizations: the Structural Engineers Association of California (SEAOC), the Applied Technology Council (ATC) and the California Universities, for Research in Earthquake Engineering. (CUREe). SEA: OC's. 4000 professional members. are responsible for designing and retrofting WSMF buildings and, through the organization's extensive committee structure for the devel- opment of professional practices and seismic design rec- ommendations. ATC is recognized for its conduct of pro: fessional investigations and the development of consensus- backed design recommendations. CUREe is comprised of the eight largest research universities in California and spe- cializes in the management of large, multiisciplinary research programs related to earthquake engineering. The Wide expertise and close proximity of these organizations permits effective management of the program. However, all aspects of the program are conducted with the active involvement of engineers, researchers, construction experts and others from throughout the US. The FEMA/SAC Steel Program is divided into two major phases. The first phase focuses on the development ‘of Interim Guidelines for the inspection, evaluation, rep modification and construction of steel structures. Tis phase is supported by limited amounts of laboratory and field test- ing, as well as other topical investigations. Major efforts to identify and verify reliable and costeffective long term solutions and to develop seismic design criteria and stan- dards of practice for ste! frame structures are contained in the second phase. Currently, Phase I has been completed and the Phase 2 efforts are underway. The tasks undertaken in Phase | are schematically illustrated in Figure 1 [Peer Aten Sia oie Survey of Baling Danae f = EE ‘Primary Experiment Prgram ‘ional nig ond aly Areas Propo Figure 1 Task flow chart for Phase 1 of the SAC Steel ProjectDamage to steel! buildings during the Northridge earthquake: S. A. Mahin 263 ‘The backbone of the FEMA/SAC Steel Program is the development of specific design advisories, guidelines and other criteria for design, inspection, evaluation, repair, ‘modification and rehabilitation of WSMF structures. The Interim Guidelines" developed in Phase | were writen by committee of 10 experts from a variety of disciplines. ‘The Guidelines were subjected to extensive review by engineers, researchers, building regulators and other public officials, and representatives from the steel and construction industries. The scope of the Interim Guidelines covers welding procedures, quality assurance, post-earthquake actions and new construction. Specific chapters cover: (a) ‘welding and metallurgy: (b) quality control and assurance; (c) visual inspection; (d) non-destructive testing; (e) class fication and implications of damage; ({) post-earthquake evaluation; (g) post-earthquake inspection; (h) postearth- quake repair and modification; and (i) new construction As shown in Figure J, various preliminary investigations and tests were carried out to support the development of the Interim Guidelines and the planning of Phase 2. The following sections describe some of these studies as they relate to the inspection, evaluation and repair of existing WSMF buildings. Additional information on these investi- gations can be found? Information on activities being tundertaken in Phase 2 is presented at the conclusion of the paper. 3. Survey of Northridge building damages Information on the actual performance of steel buildings has been scarce following the Northridge earthquake. A systematic effort has been undertaken as part of the SAC Steel Program and carried out by investigators funded by the National Institute of Standards and Technology (NIST) and FEMA. Four levels of survey were conducted. In the first, in-depth inverviews? were conducted with design engineers, building inspectors, contractors. and building officials to identify problems encountered in repair of build- ings. This identified a numiver of difficulties in identifying damaged buildings and in inspecting and repairing them. In the second type of survey, a brief questionnaire was sent to more than 200, randomiy-selected owners of steel buildings to assess their awareness of problems occurring in steel buildings, whether their building has been inspected by an engineer, and the state of damage, if any, found in their building. This prefiminary survey was used to estimate the overall scope of damage to steel buildings and to help identify geographic areas where steel buildings were dam- aged. Based on this survey” and other more detailed infor- ‘mation on the ground motion intensity and structural dam age, the Interim Guidelines recommended a detailed inspection of steel buildings where ground motions exceeded 0.2 g A third level of survey was carried out by engineers on damaged steel frames**. Detaled information was obtained on 89 buildings regarding the types and locations of dam- age observed and their strictural configuration, materials and detailing, This third survey was supplemented by even ‘more detailed surveys of damage in eight buildings selected for dynamic analyses. These detailed case study buildings incorporate more precise information on local detailing, as well as on predictions of local demands. Results of these surveys were used to further identify methods that can be used to select buildings for inspection, and the joints within a suspect buildings that should be inspected. For example, it was found that, on average, 70% of the floors of the buildings surveyed had serious damage {0 at least one welded joint. Only 25% of the connections were found with no damage. About 20% of the building frames included in the study had more than 40% of their connections damaged. In some instances, all of the connec- tions at one or more floors were damaged Survey results also were used to assess methods of pre- dicting damage. For instance, the data show a correlation between the area supported by a WSMF connection and the damage ratio. This may suggest that redundancy is a contributor to improved response (fewer columns pe pate in buildings of the same area as the column tributary area increase), but other factors related to the larger size ‘members, and later construction dates, likely with bigger tributary areas may be contributors. Ground motion inten- sity was found to be correlated to damage, but the lack of data at high peak acceleration values makes precise corre- lation difficult. In spite of this general trend, significant damage was detected at sites with low peak ground acceler- ations. Similar investigations of the damage showed that the ‘damage in low rise structures was more or less uniformly distributed over height (with slightly more severe damage occurring near the bottom of the structure), whereas tall buildings exhibited greatest damage in the upper half. This, ‘may relate to more pronounced higher mode effects in taller buildings, but other factors (e.g. relative size and strength, of beams and columns, axial load intensity in columns, ete.) ‘may also contribute. Survey results also show that damage tends to congregate. Thus, finding a severely damaged con- nection as part of an inspection process should trigger sspection of other nearby connections. Comparisons of survey results with results of elastic analyses of the buildings (using simulated Northridge earth- ‘quake records developed for the building sites* show rela- tively poor correlation. Analysis results show that the most, heavily stressed joints are most likely to be damaged; how- ever, the precise location of damage and its magnitude can- not be reliably deduced from results of conventional elastic dynamic analyses. The 60% most highly stressed of the ‘connections in a structure (relative to their capacities) all hhave roughly equal chance of being damaged. Thus, analy~ sis may not be a good approach for assessing the particular joints to be inspected, though it may be able to indicate the floors that should be included in the investigation, 4, Detailed analyses of buildings damaged in the Northridge earthquake Twelve buildings damaged by the Northridge earthquake were selected for detailed analysis by consultants using clastic and nonlinear analysis programs. To support this, effort, detailed investigations? of ground motion character- istics during the Northridge earthquake were developed. This consisted of gathering available strong motion records in or near the subject buildings. In addition, a fault disto- cation model was constructed, verified with available rec- ‘rds and used to generate time history estimates atthe sites, Of the case study buildings. In addition, time histories were generated at other heavily shaken ateas throughout the ‘greater Los Angeles area to assess the effect of location ‘on damage. Buildings were selected with heights ranging from 2 to 17 stories, and in locations varying from Santa Clarita on264 Damage to steel! buildings during the Northridge earthquake: S. A. Mahin the north of the epicenter to Santa Monica on the south, The analyses were intended to help identify the causes of the damage, as well as the reliability of various analytical methods and modeling assumptions in predicting the observed damages. For this reason, heavily damaged build- ings were excluded in the selection of the case study struc ‘ures. In two cases, buildings without any apparent damage were included: these buildings were immediately adjacent to other damaged buildings included in the study. In four of the buildings, recordings of response during the Northridge earthquake were available and in three other cases, ambient Vibration tests were performed to help benchmark the dynamic characteristies of the structures ‘The analysis results?” indicate that the case study build ings were very strong in comparison with the design forces incorporated in building codes. In many of the buildings the estimated response spectrum for the sites were nearly double those considered in current building codes (assuming elastic response, Ry=1). The elastic analysis results showed that the most heavily damaged buildings were only stressed 2-3 times their capacities; in several cases, damaged buildings were predicted to remain essen: ly in the elastic range of response, suggesting that the buildings may be 48 times stronger than required by code. ‘The main reason for this over-strength appears to be the use of large-sized members to satisfy stringent drift require- ‘ments contained in the applicable building codes. ‘The analyses also show limited to poor correlation with observed damage. The general areas of high damage could be identified, using high over-stresses or large interstory drift ratios, but a indicated previously, the severity of dam- age is not well correlated with demand, Similarly, areas of Jow computed stress were subject to damage. The reasons for these differences include the effect of initial defects and poor workmanship, and the limitations of the analytical ‘methods in modeling actual behavior. More detailed elastic and inelastic analyses shed some light on this last possiblity. First, both types of analyses indicate that higher mode effects are very important. As a result, equivalent lateral static force methods in the elastic range and nonlinear push-over analysis are of limited value for longer period structures. Similarly, the predicted distr- bution of damage in these longer period structures is highly sensitive to small perturbations in the characteristics of the ‘ground motions considered. The presence of high velocity pulses in the ground motion records also resulted in sub- stantial higher mode response contributions even for short structures, with similar high fluctuations in response with cchanges in input motions. ‘Another aspect of this difficulty is the traditional models ignore or only crudely treat deformations in and near the beam-to-column panel zones. More refined finite element analyses, especially those performed in the inelastic range, indicated that these panel zone deformations were very si nificant and disregarding their effects reduced the realism of the models, A similar observation was made regarding the importance of the slabs acting in composite or partial composite action with the floor beams. Inclusion of slab ‘effects had an important effect on global stiffness and per- jods, and local stresses and deformations. Howev. ‘methods for modeling slab and panel zone effects are quite simple and may be inadequate in many situations. ‘The conventional assumptions used in the analysis and design of beams at the face of the column (where the majority of the brittle fractures appear to have initiated) may be seriously flawed, For instance, most analytical mod- els for beams are based on an assumption that plane sec- tions remain plane during deformation. However, review of experimental data and finite element results suggests that this is far ftom true, with high local bending and shear deformations being induced in the beam and column flanges. This is due to a number of factors, but is especially pronounced when plastic shearing deformations occur in the panel zone. Similarly, the distribution of shear stress in the beam's web is not uniform in such cases, often concen: trating the majority of te shear force in the highly stressed beam flanges. ‘Actual material properties are not niform, and vary ran- domly from member to member and systematically with loading direction, section size, and welding. procedures. Normal member to member variation of material properties will result in a column that is weaker than the supported beam (eventhough the same grade sel is speced use of the transverse restraint provided by the nendcd column flange, the ot may Rave nee gible inelastic deformation capacity in this situation. Situations where the beam is significantly stronger than the column may thus fail in a brittle manner as a result of the under- lying mechanics, regardless of the quality of the base materials and weldment 5. Parametric analytical investigations ‘A series of analytical investigations were undertaken to assess the effects of ground motion characteristics, struc- tural modeling, member fracture, structural configuration were also undertaken", Estimates of plastic rotation demand for various types of WSME structures were developed and the effects of connection fracture were stud- ied. It was found that hypothetical buildings (especially shorter ones) subjected to the severe shaking representative of the lightly populated area north of the San Fernando Valley were most susceptible to collapse or severe damage. 6. Preliminary test program A total of 37 tests of full size beam-to-column connection subassemblies were done as part ofthis preliminary invest gation". Twelve test specimens (like those shown sche- matically in Figure 2) were constructed in utilizing pre- Northridge details, half of the specimens had W36 x 150 beams and half had W30%99 beams. 14-in wide-flange sections were used as columns in both cases. Dual certified (F,> SO ksi) steel was used in all cases. Slabs were not included in any of the specimens. These specimens exhil ited brittle appearing fractures; some specimens had fr tured without any plastic deformation, while others were able to deform to a plastic rotation of 0.02 prior to fractur- ing. ‘These damaged specimens were repaired or upgraded =r Figure 2. Test specimens considered in Phase 1 programDamage to stee! buildings during the Northridge earthquake: S. A. Mahin 265 Repair consisted simply of rewelding the connections using. high notch tough FCAW procedures; backing bars were removed, the root pass of the CJP weld on the beam flange to column flange connection was air-arc gouged and repaired witha fillet weld. This is the prevalent practice in repairs of damaged buildings in the Los Angeles area. T results indicated that the repaired specimens, constructed with careful quality control, were able to retain their pre- ‘damage strength and stiffness, Plastic rotation capacities were not significantly different from those achieved in the first tests. Thus, improved workmanship and materials did not significantly improve the inelastic performance of these details Some of the specimens were upgraded in an attempt to improve their plastic deformation capacity. In the cases investigated, inclined haunches were applied to one or both sides of the beam at its connection to the column, This detail attempted to move the plastic hinge in the beam away from the face of the column to the end of the haunch, These {ests supplemented earlier tests conducted at the University of Texas"? and other tests" for a summary of current test results which utilized trapezoidal and rectangular shaped cover plates, vertical fins, or side plates. Test results on the triangular haunches indicate that they are generally able to increase the plastic deformation capacity of the connection to at least a plastic rotation of 0.03, Four beam-column assemblages were removed during demolition of a heavily damaged building. These specimens ‘were tested in their damaged condition as well as following repairs. These tests were correlated to detailed elastic and inelastic analyses of the building’ Simple weldment specimens (Figure 3) were fabricated and tested to assess effects of various weld procedures, initial defects, repair methods and loading rates! These specimens represent the connection of the beam and col- tumn flanges, but do not replicate the complex stress state induced in an actual beam to column connection. These results clearly demonstrate the importance of quality weld- ing and the greater reliability that can be achieved with high notch-tough weld consumables. Tests were also conducted on a few details appropriate for new construction. These specimens utilized the same steel materials as for the previous beam to column tests, but utilized high notch tough weld wire. In addition they included reinforcement of the end region to shift the plastic hinge region away from the face of the column. These specimens utilized horizontal cover plates and horizontal hhaunches. Generally superior behavior was obtained in these tests! Figure 3 Schematic illustration of simple weldment test speci- 7. Recommendations for evaluation, inspection and repair Based on the results of the studies described above, additional studies undertaken by others, practical consider- ations and expert opinion, a methodology was developed 4s pat of the SAC Steel Program forthe inspection, evalu- ation and repair of WSMF buildings subject to damaging earthquakes. Because of the many uncertainties involved, these recommendations are considered interim and. may change significantly as more information becomes avail able, The overall process is illustrated in Table J. This con- sists of a preliminary evaluation in which key parameters are examined t0 identify the need to conduct a detailed inspection and evaluation of a particular building. A build- ing believed to have sustained structural damage is then subjected to additional analysis and physical inspection to asse8 its likely slate of damage, represented by two para- meters: D, and P, where D, isthe likely damage index at 8 floor ‘" and P'which represent the probability that the damage index for any floor in the building exceeds 4, respectively. Recovery actions 0 be taken depend on the values of D and P, and range from doing nothing to com- pletely modifying’ the structural system. Detailed. pro- ‘cedures for repairing or modifying damaged connections are included in the Interim Guidelines’. Additional infor- mation on each step is presented below. 8. Selection of buildings for detailed inspection The wide spread damage to welded and other connections in WSMF buildings during the Northridge earthquake sup- ports the need for rigorous evaluation of WSMF buildings -orporating vulnerable welded moment-resisting connec: tions to determine the extent and implications of any dam- age sustained. The Interim Guidelines identify several trig- gers that would be appropriate to initiate a detailed evaluation. These include: estimates of ground shaking in excess of 0.2 g (or 0.3 g ifthe magnitude of the earthquake is less than 6.5), location close to a fault rupture, structural fof severe architectural damage to neighboring engineered steel or reinforced concrete buildings, architectural damage for permanent drift (> 0.5% story height) in the building, ‘unusual damage during aftershocks, or a request by a build. 1g official. ‘The Interim Guidelines indicate that a structural analysis is not necessary, but may be performed. If an ensu- ing analysis indicates that the members of the building all remain well within the elastic range during the earthquake (fs 4: (7) determine appropriate actions from Da. and (8) prepare report to owner. All ofthe welded steel moment connections in the struc ture are to be divided into groups expected to have similar mechanical characteristics. It is also desirable to divide the connections into zones within the building; for example, a building might be divided into wo separate zones con- sisting of connections oriented along one axis of the build- ing, and the other zone consisting of connections oriented in the perpendicular direction. Int his manner, directivity effects, noted in the Northridge earthquake, may be accounted for in developing damage statistics and esti- mates. Similarly, if a building is subject to torsional motion, it might be divided into four zones, An initial sampling of connections from each group is to be made. Visual and non-destructive methods are used to identify structural damage in these connections. Three ‘methods have been recommended in the interim guidelines to select the connections to be inspected. These include a random selection process, in which 10-30% of the connee- tions are inspected in a group, with a larger percentage required when there are smaller numbers of connections in ‘a group. Because of the weak correlation between analytical predictions and observed damages, it was felt adequate t0 randomly select the locations of the connections to be inspected. Up to 30% of the locations of connections to be inspected using this selection procedure can be relocated to account for results of analysis, iregularities in the system, use of large size members in certain areas and practical con. siderations, {In some cases it may be desirable to use a simpler means Of selecting inspection sites. A deterministic method has been developed in which, for each frame in the structu fat Teast one connection is selected on every column face and at least one connection is selected on every floor. Li tations are placed on the number of connections that can bbe inspected at a particular floor or column line. While this, method is simpler, and provides mote flexibility, it typically results in greater numbers of connection to be inspected. ‘An analytically based selection method may be appropri- ate where detailed rational analyses have already been per~ formed on the building. In this case, the number of connec- tions to be inspected is the same as for the random procedure, but up to 60% of the inspection locations can bbe identified on the basis of the results of the rational analy. sis. The remainder of the connections are to be selected from connections distributed throughout the building, including upper, middle and lower stories. The Interim Guidelines recommend that if this analytic approach is, used, that the analysis and the locations of the connections, {o be inspected be subject to independent third party Inspections are then carried out on the selected moment resisting connections. It is desirable to rationally estimate the effect of any detected damage has on the stiffness, strength and deformability of the connections. Given the difficulties in making this assessment, a connection damage index dj is suggested in the Interim Guidelines. The connec- tion damage indices vary between 0 (for no damage) an 10 for total loss of effective capacity. Various categories of damage are considered, including damage to welds, girders, columns, shear tabs, and panel zones. Methods are sug- gested for assessing the effects of several types of damage ‘occurring in a single connection, Examples of damage indi- ces associated with girders attached to a connection are shown in Table 2 It is recommended in this procedure that the number of inspections be increased in the vicinity of any severelyDamage to steel buildings during the Northridge earthquake: S. A. Mahin 267 Table 2 Typical damage irdices 4, corresponding to girder damage, excluding damage to welds Connector damag Observed damage Index. d, Buckled flange 4 Yieided flange 1 Flange fracture (top or bottem} 8 Flange fractures (top and bettom) 10 Web yielding/buckling 0 Web fracture 10 a Lateral torsional buckling damaged connection. Thus, whenever a d, greater than 5 is, found, inspections are to be extended to the immediately adjacent connections (one connection left and right, and. fone floor above and below). Where a joint damage index greater than 9 is detected, inspection of two connections. immediately adjacent (two connections to the left and right 1s well as two floors above and below) to the damaged connection is recommended. This and other requirements, (Table 3) will substantially enlarge the number of connec tions that need to be inspected in severely damaged struc~ tures. Itis also recognized that, when severe damage occurs in a WSMF building, the structural deformations causing the damage in the welded beam-column connections may cause damage in other types of welded or bolted connec tions. Damage was observed in buildings following the Northridge earthquake in simple shear connections used as part of framing not intended to be part of the lateral load, resisting system. Similarly, welded joints resembling con- ditions present in vulnerable beam-column connections exist in other connections including connections of shear links to columns in eccentrically braced frames, column splices, column base plates, and so on. Thus, the Interim Guidelines recommend that inspection be extended to these types of connections under certain circumstances. To assess the safety and need for specific recovery actions, a floor level damage index D, is computed. This, index ranges from 0 (no damage) to 1 (complete loss of ‘connection capacity at a loor). First-order, second moment Table 3 Recommended recovery actions 9s a function of P 1d Dae Damage state Recommended recovery action All cases Repair all connections found with 628 > 5% OF Dex0.1/ Repair all connections found with Ge? Inspect all connections. Rey ‘connections found with dj 2 Inspect all connections. Unsafe conditions may exist, careful ‘assessment of building. safety is jr all connections P> 10% OF Drax > 0.2 ir all P> 28% OF Onan > 033, Dos 05 Inspect all connections. Unsafe condition exists. Careful assess ‘ment of structure. is required. Repair all damage and modity con: nections or system for improved performance. statistical procedures (assuming normal distributions of, damage) are used along with the d, values determined for the inspected connections to estimate the maximum likely value of D, at a floor and the likelihood (P) that any D, exceeds 4. The value of P and the maximum D, value for all oors (Dre) is used to determine appropriate recovery actions. Recommended recovery actions are shown in Table 3. ‘When a WSMF has experienced damage to a significant percentage of its moment-resisting connections (on the order of 25% in any direction of resistance), in addition to repair, consideration should be given to modifying the configuration of the individual damaged connections and possibly some or all of the undamaged connections to pro- vide improved performance in the future. However, partial modifications to the structural system should be made with ddue consideration of the effect on overall system behavior. When a WSMF building has had many seriously damaged connections (on the order of 50% in any direction of resistance), owners should be informed that this damage may have highlighted basic deficiencies in the existing structural system, or a geologic feature which unusually amplifies site motion. In such cases the existing system should be both repaired and modified to provide an acceptably reliable structural system, Modifications may ‘consist of either local reinforcement of individual connec tions and/or alteration of the structure's basic lateral-force- resisting system. Such modifications could include the addition of braced frames, shear walls, energy dissipation devices, base isolation and similar measures, ‘The engineer performing evaluation of the building should prepare a detailed report to the owner regarding the findings of the inspections and evaluation, along with a summary of the recommended actions. Specific recommen- dations for the content and format of the report are con- tained in the Interim Guidelines. 10. Inspection procedures ‘Since the evaluation of the building depends on the results, Of the inspection, the Interim Guidelines contain precise recommendations for the conduct of the inspections. These include methods for exposing connections (removal of architectural finishes and fireproofing), methods for visual inspections, and nondestructive testing, qualifications of inspectors, and a standardized classification and reporting, system for the damage detected. Emphasis is placed on the reed for careful visual inspection, as supplemented by the use of magnetic particle testing and liquid dye penetrant testing. While the use of ultrasonic testing is encouraged, diffi- culties exist in detecting fractures in joints, especially at the beam-column connections where backing, runout tabs and other conditions may lead to spurious images. Thus, ultrasonic testing should not be used as a substitute for other forms of inspection, 11, Repair of earthquake damages Currently, the vast majority of connections damaged by the Northridge earthquake are being simply repaired to their condition prior to the earthquake. As used in the Interim Guidelines, the term repair means restoration of the sirength, stifiness and deformation capacity of structural ‘elements that have been damaged or have constructionFigure + Back-gouge and re-weld of typicsl beam to column flange connection defects. Modification means actions taken to enhance the strength, stiffness or deformation capacity of either dam: aged or undamaged elements, or to the structure as a whole. Repair is intended to restore a building's pre-earthquake seismic resistance, but not significantly improve its original levels of safety or reduce the inherent seismic risk. In future ‘earthquakes a repaired structure will experience either more or less damage, depending on the particular characteristics ‘of the event. Structural modification (upgrading) can be used both to improve building safety and reduce seismic risk. Connections which have been damaged can be econ- comically modified at the same time repairs are made. Modi fication of connections throughout the structure, or pro- vision of an alternative lateral force resisting system, will likely substantially improve building performance; how: ever, this will entail a significant cost premium. Modifi- cation of only some connections, and not others, may cause an increase in vulnerability, due to unbalanced concen- trations of stiffness and strength. Currently, there is no clear evidence that one welding procedure can always produce better welds and structural performance. However, it appears that poor workmanship and inadequate structural details can produce poor perform- ance regardless of the welding procedure, Thus, it is imperative that repair techniques fully account for proper welding procedures and incorporate rigorous quality control and assurance programs. A written Welding Procedure Specification is required to provide specific instruction to the welder and inspector for each weld condition encoun- tered. In particular, care must be exercised in using proper pre-heat and post-heat, electrode diameters and deposition rates, and inspection procedures, Special reference is made in the Interim Guidelines to the desirability of using weld filler material with high impact toughness (> 20 ft-Ibs at OF), and the need to avoid mixing FCAW and SMAW in repair situations ‘Several examples of repairs are cited in the Interim Guidelines. Two of these are shown in Figures 4 and 5. In situations where a fracture has occurred in or near the weld New beam fange plato Figure 5 Example of beam flange repair by replacement with stool plate Damage to steel buildings during the Northridge earthquake: S. A. Mahin joining the beam and column flanges, it is recommended that the backing bar be removed and the root pass be air arc back gouged and subsequently repaired with a fillet ‘weld, Any fractures in the weld and base metal should be air-are gouged, ground to achieve proper fit up, rewelded, and inspected, In some circumstances the fracture may extend into the column (e.g. a divot of column base material may be removed by the fracture) and it may be necessary t0 rebuild the column section by filling the divot with weld material, grinding the repair weld surface smooth, and inspecting the repaired column flange. In all ‘welding situations, the engineer should consult with know!- edgeable experts in the fields of welding and inspection, In some cases it is necessary to replace portions of a ‘member. For instance, it may be necessary to remove the bottom flange of a beam due to severe local buckling or 10 gain access to make repairs to the column flange. In such ceases, the beam flange may be replaced by a plate. Care needs to be exercised in orienting the plate with the rolling direction in the proper direction, In some cases, the plate ccan be welded in the same location as the original flange (welding the plate (0 the beam web along the K Tine as well as to the column and beam flange). In other cases it may be acceptable to weld a plate to the surface of the existing flange (see Figure 5). In both cases, care must be taken regarding local details, welding procedures and avoidance of local or lateral buckling of the flange and/or web. It may be desirable to improve the behavior of a severely ‘damaged connection through modification. Currently, most approaches (see Figure 6) involve moving the plastic hinge away from the face of the column where welds join the ‘beam to the column, This may be accomplished by locally strengthening the beam near the column face, or locally weakening the beam at a location where the plastic hinge is desired. Various details have been suggested for both of these approaches (see Ref.[1] for a more complete discussion). Local strengthening can be accomplished by addition of cover plates, side plates, vertical fins, haunches ‘and knee braces. Weakening of the beam away from the face of the column can be achieved by reducing the area Of the flanges by drilling holes in the beam flanges or by ‘rimming the width of the flanges. Because behavior of modified details have not worked satisfactorily in all cases, it is necessary to confirm designs using tests or test-validated calculations. Recommendations. ‘are made! regarding the loading and instrumentation pro- ss well as for situations where new tests are required. rotation capacities recommended for modified details for use in existing buildings are 0,025 radians. This, is lower than recommended values for new buildings (a Value of 0.03 is stipulated), since it is not expected that Sits ios =a Figure 6 Typicel approaches for modifying connections to move plastic hinge away from the column faceDamage to ste! buildings during the Northridge earthquake: S. A. Mahin 269 existing buildings will be designed for the same level of reliability as new structures. Design recommendations for modified details are based ‘on realistic estimates of accual material properties including strain hardening effects, Similarly, moving the plastic hinge towards the midspan of « beam will increase the plastic rotation demands and, when the beam is strengthened at its ‘end, increase the moment at the face of the column. This inerease in moment will require careful evaluation of demands on the panel zone and column, in order to avoid unwanted plastic deformations in these locations. 12. Overview of Phase 2 efforts ‘The Phase | Interim Guidelines (FEMA 267) provide the best answers within the current state-of-knowledge on what, to do about WSMF problems. However, Phase 1 has dra- matically demonstrated the limitations of our current knowledge. The substantial damage, including collapse, experienced by many modern WSMF buildings in Kobe during the Hyogoken-Nanbu earthquake has heightened the appreciation worldwide of the importance of developing, reliable, practical and cost-effective solutions to this prob- Tem, “The Phase 2 effort addresses these issues through 11 inter-related tasks spanning over 48 months. ‘These tasks include: (a) assess past performance of steel buildings (b) assess current knowledge; (c) conduct topical investigations; (@) write state-of-the-art reports; (e) conduct test program; (f)_ develop inspection procedures; (g) draft and review Seismic Design Criteria; (h) assess economic, social and political costs; (i) finalize and submit Seismic Design Criteria; Gj). prepare code version of the Seismic Design Criteria; (k) develop training materials; (1) disseminate information ‘The Phase 2 steel project culminates in the production of four major deliverables. All products will be applicable to buildings located throughout the US, in areas of low, ‘medium and high seismicity. These deliverables include the following. 12.1. Seismic design criteria for steel moment frame To address deficiencies observed in many steel buildings following the Northridge earthquake, engineering rec- ‘ommendations are being developed for design, constru tion, and quality assurance of steel moment frame buildings related to: (a) new structures; (b) existing structures that, are to be rehabilitated prior to a potentially damaging earth- ‘quake; and (c) damaged structures to be repaired following an earthquake. The basis of the criteria contained in the document will explicitly consider performance-based design concepts. 12.2. Inspection procedures Procedures and accompanying guidance will be developed to: (a) inspect construction for new buildings and for repair ‘and rehabilitation of existing ones; (b) inspect existing, undamaged WSMF buildings in order to assess their vul- nerabilty to future seismic events; and (c) determine, fol- owing an earthquake, whether and to what extent a build- ing has suffered damage, 12.3. State-of-the-art reports Resource documents will be prepared summarizing know!- ‘edge related to steel frame design and construction, as well as the performance of steel frame structures under earth- ‘quake loads. Separate state-of-the-art report will be developed on the following general topics: (a) materials and fracture; (b) joining and inspection; (€) connection performance; (d) system performance; (e) performance prediction and evaluation; (D) performance of steel buildings in past earthquakes. 12.4. Report on the economic, social and political impacts of the Seismic Design Criteria ‘The economic, social and political costs and ramifications of implementing, or not implementing, the Seismic Design Criteria, and identifying specific means to ameliorate adverse impacts, will be assessed and described in a report. ‘These work products also to be supplemented by a wide variety of technical reports, Design Advisories and Techni- cal Bulletins. A world wide web site will be used t0 facil {ate dissemination on information, ‘The conceptual process underlying the technical approach for the Phase 2 Steel Project is schematically illustrated in Figure 7. It involves seven basic steps (1) Synthesize current and new knowledge into a form usable by the design profession, the construction and steel industries, building officials, and others, as well as by developers of design guidelines, standards of practice and building codes; Evaluate this knowledge to identify specific infor- mation needed to develop and verify the Seismic Design Criteria, and to realize the objectives stipulated for Phase 2; (3) Generate the needed knowledge through research, test- ing and design applications related to different aspects of steel construction, specifically including, but not limited to, welded steel moment frame structures, or through acquisition of research results and other infor- ‘mation generated by others; (4) Develop recommendations for inspection, evaluation, Q iol \moree" Figure 7 Conceptual process utilized in Phase 2 Steel Project270 Damage to steel buildings during the Northridge earthquake: S. A. Mahin Figure 8 Logical network diagram for base period of Phase 2 Project. (Major work products are shown in boxes with darker repair and rehabilitation of existing steel frame build- ings, and for the design and construction of new ones; (5) Evaluate the technical merit and feasibility of these recommendations, including an assessment of their economic, social and political costs and impacts ‘through peer assessments, trial designs, physical testing and other means; (6) Develop, review and finalize Design Criteria and other documentation; and (7) Implement a’ knowledge dissemination program, including the systematic development of design aids and training material ‘The intent of this process is to foster innovation, yet to develop design methodologies and details that are reliable, feasible and economic. This entails balanced consideration of the multidisciplinary technical, professional, economic, social and political issues involved. ‘To implement the conceptual process diagram shown in Figure 7, the tasks are organized according to the logical network diagram shown in Figure 8. Similar to Figure 1, this diagram shows the major work products aligned along the top, with supporting assessments, investigations, studies and research positioned below along with various monitor- ing, quality control, and information dissemination func- tions. 13. Concluding remarks While the Interim Guidelines represent current US thinking fon the proper evaluation, inspection and repair of WSMF buildings, there are clearly many uncertainties and unre- solved questions. In Phase 2 of the FEMA/SAC Stee! Pro- gram additional research and testing will more clearly define the parameters controlling the performance of con- nections and systems, and develop and verify reliable and cost effective procedures for design of new moment frame buildings and for the evaluation and rehabilitation of exist- ing ones. Phase 2 has a duration of 4 years and efforts will ‘be made to cooperate with activities undertaken worldwide ‘on this topic. Acknowledgements ‘The information reported in this paper was prepared under ‘contract (EMW-95-K-4672) to the US Federal Emergency Management Agency as part of its Program to Reduce Earthquake Hazards in Steel Moment Frame Structures. ‘The assistance and support of Michael Mahoney and Robert Hanson, FEMA’s Program Officer and Technical Advisor, respectively, are greatly appreciated. One of the authors of this paper is the Program Manager, for the SAC Steel Pro- gram. The authors are indebted to the efforts and work of ‘numerous individuals and organizations that have supported ‘of contributed to the SAC Stee! Program, especially 10 Ron- ald Hamburger, Project Director for Product Development and James Malley, Project Director for Topical Investi- gations. The findings, conclusions and recommendations summarized in this paper are not intended for use in design or evaluation of structures. They do not represent official policy of FEMA or the SAC Joint Venture References 1 SAC Joint Venture, “aterm guidelines: evaluation, repai, mali ‘ation and design of welded steel momen frame sractres- Report FEMA 267, Federal Emergency Management Apency, Washington DC, 1995, 215 pp 2 Gates, W. and Morden, M, “Lessons frm ingpetion, evaluation, repair and constuction of welded sce momeat frames Surveys and sscessment of damage to buildings affected by the Northridge eat ‘huske of 17 January 1994, Technical Report SAC.95-06, SAC Joint Venture, Sacramento, CA, 1995 3. Durkin, ME "Inpecton, damage and repair of sel frame bul ings folowing the Norridge earthquale’: Surveys and asessnen. ‘of damage to buildings afected by the Norridge earthyuake of 7 January 1994, Techical Repon SAC 95-06, SAC Joint Venture, Sac ramento, CA, 1995, 44 Bonowitz,D! and Youssef, N. "SAC survey of sel moment frames fected by the 1994 Northridge earthquake’ Surveys and asessnes, ‘of damage to buildings affected by the Norridge earthquake of 17 Sanuary 1994, Techical Repon SAC 95-06, SAC Joint Venture, Sac ramento, CA, 1995, 5 Yousef N, Bonowit, D. and Gros, J."A survey of steel moment resisting fame buildings affected by the 1994 Nondvidge car ‘quake’, NISTR-5625, Gaithersburg. MD, 1995 6 Somenill,P. eta. "Charstrization of ground motion ding the orhridge eaquake of 17 January 1993), Technieal Report SAC 95.03, SAC Joint Venture, Sacramento, CA, 1995, 179 pp 7 eawinkler, H. Systems analysis and behavior’. Design applications ‘Analytical and eld investigations of buildings affected by the [Northridge earthquake of 17 January 1994", Technical Repor SAC {95.0 (pars A and B). SAC Joint Venture, Sacramento, CA, 1995 300 pp 9 “Case xodies of steel moment frame buiking preformance in the [Norge earthuake of 7 January 1994", Technical Report SAC 95107, SAC Joint Venture, Sacramento, CA, 1995, 260 pp 10. ‘Parametric analytical investigations of ground motion and structural response, Norvidge earthquake of 17 January 1994", Technical Report SAC 95405, SAC Joint Ventre, Sacramento, CA, 1995, Zhe 11 “Experiment investigations of beam to column connections, Tech- ried! Report SAC 96.0 (Pans A and B), SAC Join Ventre, Sara mento, CA, 850 pp 12 Engethart,M. a Sabol, “Testing of welded sts! moment con rections in response 10 th Norge earthquake’, Progress Report, [AISC Advisory Sub-commitee on Special Moment Resting Frame Research, 1994 13. "Test sutumary repons", Techical Report SAC 96-02, SAC Joint ‘Venture, Sacramento, CA, 1996, 140 pp 14 Kaufman, F.and Fisher, 1.A sady of the effects of materials an ‘welding factors on moment frame Weld joit performance using 2 Srallscale lesion specimen’, Technical Report SAC 95.08, SAC Sit Venture, Sacramento, CA, 1995